Guest lecture: sensing in smart lighting systems (Dr. Vana Jelicic, Tridonic)

In our Embedded Sensing Systems course (spring 2025), we welcomed Dr. Vana Jelicic, Domain Expert Sensors at Tridonic (part of the Zumtobel Group), for a guest lecture on 19 May 2025, titled “Sensing in smart lighting systems.”

The talk. Dr. Jelicic walked the students through how modern lighting turns into a sensing platform, following the path from a raw sensor reading to something useful: environment, sensor, signal, data, information, user. She compared the sensor modalities used to detect people and occupancy, from simple PIR motion sensors to color, thermal, time-of-flight, and vision-based sensors, each with its own trade-offs in cost, privacy, and what it can actually tell you. A recurring theme was getting more out of simple, low-cost devices: turning a PIR’s analog signal into features and running machine learning on the edge (including TinyML with Edge Impulse), and combining networks of cheap sensors for tasks like occupancy counting, dwell-time measurement, and trajectory reconstruction.

Why it fits the course. This maps almost one to one onto Embedded Sensing Systems, where students study exactly this chain: transducers and signal conditioning, on-device processing, and machine learning under tight power, cost, and compute budgets. Hearing an industry domain expert frame those same trade-offs in a real, deployed-at-scale product context gave the material a concrete counterpart, and showed students where the theory lands in practice.

Why the partnership matters. Tridonic is a leading lighting-technology company headquartered in Dornbirn, and Dr. Jelicic also teaches at the FHV (University of Applied Sciences Vorarlberg). Bringing practitioners from Vorarlberg’s industry into the classroom is exactly the kind of tie we want to strengthen at ICV-HSG: it grounds teaching in real deployments, opens up student projects and theses, and keeps knowledge and talent circulating in the region.

Looking ahead. The course will keep its core content but is being reframed as Machine Perception and Communication, carrying sensing systems further into the world of AI and robotics. Guest lectures like this one, connecting embedded sensing to real-world perception, are exactly the bridge into that direction.

She closed by inviting collaborations, a note we happily echo. Many thanks to Dr. Vana Jelicic and Tridonic for a generous and engaging lecture.